Your example has valid points, but explains a short period of time. However, Daco-Romans, Free Dacians and Slavic people coexisted in the same geographical region since at least 6th Century CE ( most historians agree). Back then, there were no borders and people were kind of trying to understand each other with the languages they had (especially without written language, they were borrowing each other words). I see this process covering a long period of time, so the language of Romanians will have more Slavic words if the Slavic DNA (hence more participation to the gene pool from the Slavic side) will be at higher concentration.
I'm imagining the Slavic movement towards South-South West kind of like a slow river (bare with me on this one). We know that water absorbs minerals from the ground as it flows, right? And let's imagine that the original Slavic people had predominant R1A yDNA, and Dacian/Thracian people had predominant I2a-DIN. Moving south they mixed with people that they met. First, the Free Dacians (that were living outside what used to be Roman Dacia), then the Romanized Dacians( already mixed with other DNA from the Roman occupation) and then the Thracian people all the way to Serbia, Croatia and so forth. Now, based what I saw as far as the I2a-DIN map: Moldavians, have more I2a-DIN than Romanians (because of more Free Dacian contribution to their DNA, and we presupposed I2a-DIN being Dacian). Moving south-southwest mixing of the people continue over a longer period of time that we think and the predominance of R1A yDNA gets more diminished, and by the time it got to the what know is Serbia-Croatia region, I2a-DIN predominance is growing stronger (higher concentration of I2a-DIN in that region, right?).
It might be oversimplistic, but it kind of makes sense, don't you think?